This article … why in the world would you even consider a new design doing this? Much less call it “opening yourself up to so many options for your network design and configuration”?
This is an ancient design that you still see here and there on old mainframe and “big iron” type boxes. It made sense then for why it’s there. Now a day those reasons are pretty pointless given the technology we have on the host and the upstream router(s)/switch(s).
This article advocates defeating the point of a default gateway for the host. Beats me why in the world you would even consider that alone. It’s a horrible design philosophy and I’d love to see an actual implementation get properly drawn out on a diagram. I’d also love to see the troubleshooting workflows you would have to go through with this now that you have a dynamic state the host can be in for traffic patterns.
It’s also a nice little security hole that coudl be pretty easily exploited … we might as well flip ICMP redirects on while we’re at it.
FYI some of the dynamic routing functionality in Windows is pulled in the LH (2008) release as nobody uses it.
You must be logged in to reply to this topic.
Create a free account today to participate in forum conversations, comment on posts and more.